help me understand

[Dowadudda]

I am attempting to learn more about chillers.

evap approach, is difference between saturated suction and leaving chilled water temp correct?

As a rack guy, my question first to ask after getting approach in my head is, where is my superheat? In my head I am, thinking if I am at saturation out of the barrel I am not boiling much gas off before I hit the compressor. what am I not understanding?

[bigtime]

http://hvac-talk.com/vbb/showthread.php?t=176465

I am not a chiller guy, but I beleive some chillers have flooded evaporators. In this case there will be no superheat. Superheat robs efficiency by its nature, and is only neccasary to prevent floodback. Saturated vapor is denser, so the compressor moves more #/hr, does more frigerating.

[nyrb]

Working on DX systems verses flooded evaporators is certainly a little different. What you said is true, with a flooded evaporator on a centrifugal chiller there is little or no superheat present in the suction gas. On a DX system you would likely check your superheat as a partial measure of effective cooling. In a chiller since there is almost no superheat the best measure of cooling efficiency is to compare leaving liquid temperature to saturated refrigerant temperature, the closer the 2 temperatures are the better. Newer chillers with high efficiency tube bundles may run with 1 or 2 degree approaches at full load where older less efficient chillers may have 10 degrees or more approach temperature. Fouled tubes or low refrigerant charges generally result in higher approach temperatures and are a good indication of these conditions. Additionally most flooded evap bundles are equipped with eliminators above the tubes like a cooling tower to reduce wet refrigerant from reaching the compressor.

On a similar note condenser efficiency is also measured by approach temperatures, high approach temperatures on the condenser can indicate fouled tubes or low water flow amongst other things.

Some manufacturers pay more attention to discharge superheat to ensure chiller has an adequate charge and is performing properly. It all depends on the beast your looking at. I'm certain others here could give a much better explanation than me.

[Dowadudda]

condenser approach makes perfect sense to me. Saturated temp, subtracting leaving liquid refrigerant temp is good subcooling. The attmept to get liquid refrigerant temp close to leaving water temp allows for that subcooling (approach).

the evap approach throws me off some understanding of how superheat works itself to keep a compressor from flooding.

[nyrb]

Keep in mind that the chiller manufacturer designs the evaporator to run at a certain liquid level with some tubes wet and some tubes dry, 1/3 could be wet, 1/3 boiling foam and 1/3 dry for example. The boiling refrigerant has to travel past a number of dry tubes for lack of a better term and gains some heat as it goes by those tubes. Each manufacturer has there own means of maintaining the correct amount of refrigerant in the evaporator to prevent liquid carryover into the compressor be it low side / high side float, fixed orifice or TXV/EXV.

[klove]

I am attempting to learn more about chillers.

evap approach, is difference between saturated suction and leaving chilled water temp correct?

As a rack guy, my question first to ask after getting approach in my head is, where is my superheat? In my head I am, thinking if I am at saturation out of the barrel I am not boiling much gas off before I hit the compressor. what am I not understanding?

You're boiling XXX amount of refgt for XXX amount of refrigeration effect, whether flooded or dx. Difference is on flooded, the evap tubes are covered by what is called ebulating liquid (not pure liquid, but boiling). Normally on a flooded barrel you'll get 1 or 2 degrees suction superheat. You're doing all the latent refrigerating you can, just not superheating the suction gas to amount to much because superheat reduces efficiency. Only real way to superheat is to uncover the tubes.

You are correct on how to calculate evap approach. Saturated condenser approach is calculated by subtracting the leaving water temp from the saturated condenser temp. Subcooling is accomplished by either getting a liquid seal on the condenser outlet and running that liquid thru a subcooler bundle, driving the liquid temp down towards entering water temp (York package flooded chillers), or by using a flash economizer (Trane and Carrier) which is actually not a true subcooler, but does reduce flash gas in the evap at the point of expansion. Both ways work. Some use both (York OM chillers).

[GT Jets]

condenser approach makes perfect sense to me. Saturated temp, subtracting leaving liquid refrigerant temp is good subcooling. The attmept to get liquid refrigerant temp close to leaving water temp allows for that subcooling (approach).

the evap approach throws me off some understanding of how superheat works itself to keep a compressor from flooding.


To really answer your question, I think it would be prudent to pick a machine type.

Centrifugals, Rotary/Screws, Recips, all have there own idiosyncrasies.

If we were to pick say a centrifugal for example.
Your comment on condenser approach is incorrect, but close...

Condenser approach is really not a measure of subcooling per say, it effects the amount of cooling the refrigerant has obtained, but subcooling is a little used term on a centrifugal machine.

Condenser approach is simply a measurement of how close the refrigerant saturated temp is to the leaving water, you got that correct, but the water will be cooler than the refrigerant, the amount depends on condenser load and design of the system (not to mention condition of the tubes, refrigerant charge and compressor condition)...

For the most part, the higher the condenser approach, above design that is, the more fouled the tubes are, or the lighter on refrigerant charge it is...Subcooling would indicate a stacking issue which is possible, but rare, which is why I mentioned it earlier...

I could drag on for quite a while, so it would be easier to answer specific question, but for the most part again, the evaporator works in exactly the opposite way...

Some rotary compressors do use superheat as a way for control of refrigerant, but it is calculating discharge superheat, the approaches are figured the same way, but more ways to get handed your own butt...



I hope this was helpful...

GT

[jayguy]

suction superheat provides no practical cooling, potentially overheats the compressor and lowers the overall efficiency.

the only reason for suction superheat (or the lack of it) is compressor design.

on scroll/recip compressors, the suction gas cools the motor and any left over liquid refrigerant ends up in the crankcase diluting the oil...bad. any liquid that does not end up in the oil gets carried up and away and the compressor attempts to compress it...and we know that it bad.

by contrast, a centrifugal compressor takes the refrigerant directly into the impeller, which does not compress the refrigerant, but it adds energy by spinning it outward. the motor is typically cooled with water jackets or it takes liquid refrigerant and boils it around the winding cooling it directly. the oil is brought back into a sump...in most chillers...these are all separate items.

in screw compressors, liquid refrigerant will not usually do much harm if any at all. a screw compessor almost moves more liquid oil through it than it does refrigerant so some (lots :eek2:) liquid refrigerant is not damaging.

and approach temperatures are useful on dx systems as well...they are higher of course, however, they can be very helpful in diagnosis.

[Dowadudda]

I hope you guys understood how I was trying to relate the condenser approach measurement and how it translates to subcooling.

Leaving water temp and leaving refrigerant liquid temp.

The leaving liquid temp minus saturated temp. There is subcooling there. So translated in my brain, the condenser approach makes sense. The rejection of heat from the refrigerant to the water, from saturation to liquid temp.

Where I get lost is superheat and evap approach. I have never been exposed to the idea that a lack of superheat is lost efficiency. High superheat yes, but not a lack of it being inefficient. Saturation of refrigerant temp, where we begin to absorb heat to the refrigerant, and heat further picked up is superheat.

In the definition of superheat, it's really a measurement of heat picked up by the vapor. So. Understanding that. I am confused about how with little superheat the chiller is ideally performing with such a low superheat out of the barrel. Further, the compressor(s) on a packaged machine is close coupled and so, what is usually deployed to safely prevent flood back.

But I also see what you guys are saying. Latent Heat of Vaporization. So basically, you guys are saying the bulk of the heat absorption is being done right at saturation. So with a flooded evap, and the entire evap right near saturation, were gonna get the whole evap to absorb heat on a constant basis. The idea is to get the work done at latent. Then I see where the evap is much more efficient than stuff I work on these days.

Making sense guys.

I am trying to learn because I want to start working on these.

[ckartson]

Please keep going klove, this thread is helping me out. What do you mean by "stacking" issue? I'm getting close to understanding what you guys are saying. I'll probably have to re read this several times. What is the importance of approach data?

[retube]

Please keep going klove, this thread is helping me out. What do you mean by "stacking" issue? I'm getting close to understanding what you guys are saying. I'll probably have to re read this several times. What is the importance of approach data?

If you have running approach data for a particular machine this data can give you a good indication of the heat transfer available from the tube bundles. The flip side of this would also be if you have a low approach, your bundles are spotless and no oil logging is present maybe the compressor is lacking. Approach is also very useful in determining how hard you are pushing a machine when trying to achieve top performance. Approach is another diagnostic tool.

[klove]

Please keep going klove, this thread is helping me out. What do you mean by "stacking" issue? I'm getting close to understanding what you guys are saying. I'll probably have to re read this several times. What is the importance of approach data?

Not a lot of spare time at the moment, so we can see where this goes and get into it more later, but I'm not the one that said stacking refgt is detrimental. The fact is that actual "stacking" (building up unwanted liquid levels) of refgt in the condenser is detrimental to operation and efficiency in a condenser that is not designed to do subcooling. However, many shell and tube condensers are designed nowadays with subcooling in mind so "stacking" (if you want to call it that) is exactly what has to be done to reach rated efficiencies. I prefer to call it maintaining a liquid seal on the subcooler section.

[nyrb]

I figured if KLOVE got on this topic it would get more interesting, makes me want to eat most of my words once you see who well he explains it. I agree, keep on going as I always enjoy a good read.

[GT Jets]

Not a lot of spare time at the moment, so we can see where this goes and get into it more later, but I'm not the one that said stacking refgt is detrimental. The fact is that actual "stacking" (building up unwanted liquid levels) of refgt in the condenser is detrimental to operation and efficiency in a condenser that is not designed to do subcooling. However, many shell and tube condensers are designed nowadays with subcooling in mind so "stacking" (if you want to call it that) is exactly what has to be done to reach rated efficiencies. I prefer to call it maintaining a liquid seal on the subcooler section.

You are right, stacking is probably kind of a crude term, but some machines, like anything with an EXV for example, don't react well with a majority of the refrigerant in the condenser under certain conditions...

This is why I recommended we try to do this one type of machine at a time, or at least clarify what we are using as an example...

Recently had a Trane RTHB with a low refrigerant temp problem, stacking an overcharge in the condenser was mostly, but not entirely to blame...

Sorry if I confused anyone...

GT

[klove]

After rereading the last of 4D's posts, I would agree that we need to concentrate on either flooded or dx. Nothing was stipulated, and there has been some crossover discussion, which will get confusing in trying to keep the ideas on the same line. So Dowa, what's your preference - centrifugal/screw (flooded), or screw/scroll/recip (dx)? We can do both, obviously, but it'll be a lot easier to get them one at a time, and it all hinges on what type equipment you're wanting to get into first, since you're the OP.

[Dowadudda]

lets start with dx since I think I'll be able to follow easier, then once Igot that, I think then it will be easier to follow the flooded centrifical side. I will be able to distinguish them both in chiller language and so I don't try to compare to what I already understand.

And thanks for allowing me in here. I appreciate it.

[klove]

lets start with dx since I think I'll be able to follow easier, then once Igot that, I think then it will be easier to follow the flooded centrifical side. I will be able to distinguish them both in chiller language and so I don't try to compare to what I already understand.

And thanks for allowing me in here. I appreciate it.

Allowing you in??? Hey - you paid your money, you can dance anywhere you want. :D

Basic dx chillers with txv's for refrigerant controls are remarkably the same as what you work with already, only much simpler. All machines have a design flow rate on both barrels for water cooled, and a design ambient for air cooled. Don't ever let anyone tell you that chillers are rated at 10* delta T, 10 psid, and 45* leaving on the chilled water side cause it just ain't so!! You can swag at flow rates and temps, but the only way to know what you have is to get the design specs from the mfr and work off those (of course, that's normally a pie-in-the-sky dream, but who's lookin'?).

General operating conditions will put you at 12* to 15* subcooling on the condenser. Head pressure control will be per the chillers internal controls on air cooled, and can be for water cooled depending on mfr.

Evap operating conditions get a little subjective depending on who you're talking with. Some folks live and die with 10* superheat because somebody that "knew what they were talkin' about" told them to do it that way. What you mentioned in an earlier post is the key to understanding why you should set superheat higher - close coupled is the term. 10* sh is great for evap superheat on systems with long lines, but will provide you with tons of work overhauling compressors if you use it on package chillers. I do my best to not let the sh go below 15* unloaded, and then check it stable and loaded. Had them go as high as 25*, most hang at around 19* to 21*. Because of basic design discrepancies, this will cut down slightly on loaded capacity, but you'll save your customer a lot of compressor work due to 2 phase refrigerant flow and wet gas/flooding.

Approach temps are important on dx machines, but generally aren't looked at near as hard. They're also more subjective in this arena than in flooded eqpt because you are dealing with higher superheats, which means that you deal with actual measurable pressure drop of a flowing gas through the tubes in the evap (if you could measure it). There's just a lot to throw into that nutroll. Just as some general numbers to look at, you might normally see 6* to 12* evap approach (lvg water - sst), with superheat being provided by the higher temp of the entering water.

Everything before is a brief overview of your run-of-the-mill package dx chiller. Different ones have different idiosyncracies, so take it with a grain of salt. Then you get into the new-fangled stuff with electronic expansion valves working at 4* sh, refgt-to-refgt plate subcoolers, drain-and-fill valves instead of expansion valves, vsd's on screws, hybrid evaps like Tranes "falling film" technology, glycol instead of water, and the list goes on...(and on, and on.....).

I've done some rack work on occasion, and I can tell you unequivocally that you're wadin' in shallow water with basic chillers in comparison to large parallel systems. 'Course, they do get more complicated............

[Dowadudda]

so okay.

DX chillers straight forward. And along the same lines of what I know already. I understand the idea behind having to digest a system on site, because it may have differing components making it. A plate to plate water cooled condenser, running with a screw pump/dx barrel. Ect. And so measurement/metrics require knowing what I got and knowing the design intent of those components. So my evaporator approach temperature measurement is different on dx versus flooded. And I understand now flooded.

[pecmsg]

Maybe this will help

[Dowadudda]

to me the chiller seems quite simple. I just read that entire PDF and I am just thinking to myself, it can not be that simple.

Terminology is my largest enemy. I know how refrigerant acts. And I understand things like surge, High Side Float, ect ect.

It's just kinda neat to be able to reach back into my brain, use what I already know to teach myself something new.

I bet I could go on a call on one of these and figure it out. I'd need the controller info but. I bet I could.